Microsoft has announced that it is bringing artificial intelligence (AI) models from OpenAI to US government agencies via its Azure cloud service.

In a blog post, the US software giant said it has enabled access to OpenAI’s GPT-3 and GPT-4 models on Azure Government, which offers cloud solutions to the US government.

Microsoft has invested billions in OpenAI, the creator of the popular chatbot ChatGPT.

Since the public release of ChatGPT in November 2022, interest in large language models (LLMs) has soared.

LLMs are trained on vast amounts of internet data and generate humanlike responses when prompted.

Through Azure OpenAI Service, government clients will be able to adapt these models to suit their requirements such as content generation, language-to-code translation and summarisation, the company said.

NASA, the Energy Department, and the Defence Department are among the US federal government clients of Azure Government, according to Bloomberg.

The Defence Department’s Defence Technical Information Centre (DTIC), which specialises in gathering and disseminating military information, will be testing the OpenAI models through Microsoft’s new service, the publication said, citing a DTIC official.

Microsoft already provides OpenAI models to its business clients, and in recent months, the Azure OpenAI service is said to have experienced significant growth.

Volvo, Ikea, Mercedes-Benz Group, and Shell are among the companies using the AI service.

The latest announcement by Microsoft also marks the first known effort by a major business to make the chatbot technology available to governments.